By Doug Hecox
A little more than a century ago, Wyoming was the frontier—an untamed horizon dotted with rough-and-tumble characters struggling against the harsh and unforgiving wilderness.
Cell phones, Wi-Fi, mocha cappuccinos and other symbols of modern living are now widely available, but Wyoming remains an outpost in a new frontier: artificial intelligence.
Jeff Clune, director of the University of Wyoming’s Evolving Artificial Intelligence Lab, is paving the way toward a tomorrow in which robots will not only be able to think for themselves but also reason, adapt and evolve—much in the same way humans’ brains improved over millennia.
“Evolution is an amazing designer,” says Clune, an assistant professor of computer science for the College of Engineering and Applied Science. “If we can produce robots as smart and agile as jaguars, hawks and dolphins, we can provide tremendous benefits to society. For example, after a building collapses, we could send in robots to find survivors. If a fire is raging in the wilderness, we could send robots to put it out instead of risking human lives. The sky is the limit in terms of the benefits such robots can provide.”
The evolutionary process, Clune says, can be used to improve computers, robots and even software. “Evolutionary algorithms routinely produce designs better than human engineers,” he says.
As the 19th century’s industrial revolution led the world into the 20th century’s space age and then the computer age, Clune says more revolutions are inevitable. “The current crop of UW students will live to see the robotics revolution,” he adds. “Robotics and artificial intelligence . will continue to create new industries and produce untold benefits for society. It is great that UW is forward-looking and is investing in one of the major forces that will shape the economy of tomorrow.”
Clune’s lab relies on students with a wide array of backgrounds. From undergraduates to doctoral candidates, Clune’s crew consists of students from around Wyoming and from as far away as Ethiopia, the Netherlands and Vietnam.
“Though it’s an extreme simplification of the work we do in the lab: I build robot brains,” says Lucas Helms, a graduate student from Cheyenne, Wyo. Helms, who has been part of Clune’s team for nearly two years, earned his master’s degree in computer science in December 2014.
“We treat the brain like a mathematically defined machine that we can simulate on powerful computers to help answer fundamental questions about life,” he says. “Part of growing up in the ’90s and the new millennium was watching all the science fiction I was raised on being turned into reality. My interests in artificial intelligence and robotics are a natural extension of that.”
Doctoral candidate Anh “Totti” Nguyen of Hanoi, Vietnam, who has been part of the team since last June, agrees. “The ‘Evolving AI’ lab is the reason that I decided to pursue my Ph.D. at UW,” he says. “The research is exciting to me because I like playing with robots—and inventing new algorithms to make them smarter is super cool.”
Each Wednesday from 4–6 p.m. during the school year, Clune and his team teach local middle- and high school students about robotic engineering and software in the Laramie Robotics Club he co-founded in 2013. The meetings are free and open to the public. “From mathematics and science, to teamwork and troubleshooting, the students learn skills that will be very important for the rest of their lives,” he says.
“There is a great quote from Richard Feynman that says, ‘What I cannot create, I do
not understand,’ ” Clune says. “The same is true for our own intelligence. We learn
how it works by trying to recreate it. So the process of making artificially intelligent
robots is also a process of self-discovery, where we increasingly understand ourselves
better.”
